Perkins and BOI-485 - Testimony to Confabulation?
an Opinion by Martin Shough

It is easily possible to suppose that a letter from a stranger describing a 12-year-old memory of a dramatic event never-before heard of is a tissue of fantasy and exaggeration. When the Condon Committee received such a letter from one Forrest Perkins in February 1968 this was exactly Condon's reaction. It was, he said at the time, typical of 'the sort of rubbish we get from people' [Thayer 1987].

It is somewhat less easy to suppose that USAF intelligence officers sent a classified, encrypted report to HQ Directorate of Intelligence, Washington DC and elsewhere within two days of the event, having made up half of it out of thin air. And if they did, how explain the fact that its contents happened to closely corroborate the statement of a retired radar controller made when BOI-485 was at that time still classified 'secret' and before its existence was even suspected?

But there is a point of view that it is not proper to see this as corroboration, because BOI-485 would obviously be a record of whatever misperception was imagined by overexcited radar operators at the time, faithfully copied down by investigators over the next day or two.

It seems to me this is an improper evidence procedure, analogous to applying probabilities retrospectively and equally invalid. One should go back to 1968 and ask oneself what 'predictions' were implied by alternative interpretations of the original evidence, then judge which were fulfilled and which were not.

Perkins' first-hand account is the primary evidence source. One asks oneself, 'Is it reliable?' Condon, who was deeply sceptical, regarded it as 'rubbish'; others in the project such as Craig and Thayer did not. Condon's interpretation, could we go back and ask him, would without doubt have been that witness memory is wont to distort and exaggerate, that it would be understandable after 12 years, that therefore the account should not be taken as reliable, and that if there were original documentation available then it would probably prove far less interesting. This is in principle a very reasonable position to take and can be justified by pointing to very many cases in which exactly this sort of demystification does occur.

When BOI-485 appears, however, and when Perkins' story is found to be accurately reproduced therein, is it reasonable to say that, having already judged Perkins to be unreliable, we should therefore regard any evidence which appears to corroborate him as testimony to confabulation? Or is it more reasonable to revisit our previous hypothesis in the light of this independent corroboration?

Clearly one ought to conclude that the hypothesis of distorted witness memory is falsified, and withdraw that objection. He is accurately remembering what people believed to be true at the time. So, one may then retreat to a fall-back position and say: 'I agree that Perkins' memory seems to be excellent after all, but his judgement must have been poor then, because he is just reporting what everyone imagined to be true at the time and what they imagined to be true at the time was in fact false.' In this way BOI-485 becomes evidence of just how seductive and impressive the original misperception must have been, and only reinforces our scepticism.

But this is not a proposition that has at all the same sort of force as the original claim. Now, instead of a very reasonable distortion introduced over the years we are hypothesising a fairly dramatic misperception in real time involving a number of witnesses. This must be so, because other personnel interviewed and their written logs agreed. The tail-chase can't have been a wild delusion of one man. A number of people were involved over a length of time. Perkins even said himself that the moment when the target fell in behind the Venom was reported to him by 'the other controllers' - he was busy and missed it. He wasn't fantasising that - if anyone was, someone else was. This is consistent with BOI-485 which gives the same account of the 'tail chase' taken from the RATCC logs and interviews with personnel other than Perkins.

So we are talking multiple delusion, sufficiently concrete and fixed in the imagination of all personnel involved that a distillation of their stories recorded by third parties shortly afterwards was an accurate copy of the same 'false memory' which was exactly preserved in the memory of their Supervisor for 12 years. The argument suggests that the corroboration of independent documentary and witness sources can only be evidence of the strength of a common delusion. But there is no strong internal evidence for such a conclusion.

The existence of errors and omissions in BOI-485 does not to my mind constitute strong internal grounds. Numerical inconsistencies and typos, and a regrettable narrative incoherency inate to the reporting format, do not remove or diminish the fact that the 'core' tail-chase episode is a strong structural feature of the BOI-485 account, therein twice repeated, and independently of Perkins' account. Unless there a strong independent grounds for questioning it, it is only sensible historical method to continue to take seriously a contemporary documentary account of what was seen on the RATCC radar screens when that account is independently corroborated by the only known eye witness.

Secondary tests of this witness's reliability can now be applied. For example, Perkins' 1968 statement that the RATCC radar was a Bendix CPS-5 is found to be confirmed by BOI-485. However Perkins also recalled that there had been an extensive field modification made to the radar during which they had had to resort to the GCA CPN-4 for several days. This had been necessary, he said, because the CPS-5, originally designed [1946] for Aircraft Control and Warning work, was "lousy" for ATC purposes. One of the changes made by Bendix technical reps and local maintenance technicians had been to speed up the scan to the unusual rate of 6.66 rpm, or one rotation every 9 seconds, to improve its ATC performance.

Philip Klass [Klass 1974] vigorously challenged this claim: the standard CPS-5 scan rate given in the Technical Order applicable as at August 1956, he pointed out, was only 4 rpm. Klass argued that the new gear-train and electronic modifications required would be a "major engineering job" of a kind forbidden by the USAF since it "obsoletes spare parts in logistics supply channels". At only 4 rpm the kinds of rapid target manoeuvres reported by Perkins would have been much more difficult for the radar to "see", and Klass suggested that Perkins was improvising to cover flaws in his memory.

Perkins also stated that the CPS-5 had a facility known as sector-scanning fitted as standard (although he believed it had not been in use at the time). Sector-scanning automatically reverses the antenna turning gear at the limits of a preset arc and might in principle have allowed more frequent 'looks' at the target. Perkins was claiming that the mechanism had been disconnected anyway, but Klass's objection was that the CPS-5 manual supplied to him by Bendix indicated that the CPS-5 had never had a sector scan capability in the first place. Klass also pointed out that Perkins' mistakenly called it 'vector scanning'.

These points appeared to Klass to cast doubt on Perkins' reliability, and David Thayer agreed that the scan-rate issue in particular would, if unresolved, have a bearing on the credibility of Perkins' recollection. However, close study of the CPS-5 manual by Thayer subsequently revealed that the radar did indeed have a sector-scan capability, as well as position control or "searchlight" mode. Perkins recollection that this useful-sounding facility was disused or disconnected is also very plausible. According to Bruce T. Neale of Marconi, 'sector scan was rarely used as the wear and tear on the turning gear was considerable. There was little merit in it anyway, the time used to decelerate and accelerate the large [antenna] mass was a considerable portion of the rotation time. It was more efficient to leave it in continuous rotation . . . '

Thayer also found that the scan rate was not fixed at 4 rpm but was operator-variable up to 6 rpm, meaning that a 10 second rotation period could be selected by means of a knob on the control panel without any modification at all. Re-examination of the technical specifications then revealed that a faster rate could be easily achieved by adjusting a potentiometer in the antenna drive control circuit. There was no major engineering involved: The reason for the 9 second rotation period was simply that this was as far as the potentiometer would go.

So what was the the major field modification recalled by Perkins? BOI-485 confirms Perkins' memory that MTI was used: ". . . logs of RATCC lend reality . . . the controllers are experienced and technical skills were used in attempts to determine just what the objects were. When the target would stop on the scope, the MTI was used. However, the target would still appear on the scope." But although the CPS-5 was commonly used with MTI, the Technical Order does not indicate a delay line factory-fitted in any revision through 1957, Thayer pointed out, suggesting that the field modification Perkins remembered lasting several days was the retrofitting of the MTI unit and that this probably held up the 'official' dedication of the RATCC. This would be consistent with the fact that an August edition of the Lakenheath Base Newspaper describes the new RATCC facility as being "only a few days old".

Nevertheless it is perfectly reasonable to speculate that Perkins' account is wrong, despite the weakness of any evidence for his unreliability, and despite the appearance of corroboration in BOI-485, for reasons as yet undetermined. The issue at this point rests again on the question of prediction. We are predicting that pertinent information presently unknown would, if available, provide strong independent grounds for questioning the account. Coming up to the present, has this prediction been tested, and has it been verified or refuted?

This of course is the point at which a range of recent evidence comes into play. This evidence includes the discovery of the two 23 Squadron Venom scrambles (Brady/Chambers and Logan/Fraser-Ker) at 0200 and 0240Z, the aborted scramble of another 23 Squadron Venom (Scofield/Arthur) at 2120Z, the interception of a high-altitude balloon-like unknown by a Venom piloted by Flying Officer Dell at 1830Z, and finally the vectoring onto a radar UFO of another Venom flown by the 23 Squadron CO (Davis/McIlwrath) at an unknown time.

In addition we have the testimony of the Eastern Sector Controller, Wing Commander Grocott, that any available RAF or USAF aircraft might be called on opportunistically in such a situation by Sector or even by a GCI (as indeed is now known to have happened with the 23 Squadron CO), supporting Scofield's suggestion that an American fighter might have been involved in the first instance. There is the well-established fact that the Brady/Chambers and Logan/Fraser-Ker scrambles were neither of them under RAF GCI control, and so proveably have nothing to do with the single-aircraft interception controlled by Wimbledon's interception team at Neatishead, whilst the latter closely resembles the incident described by Perkins and BOI-485 also involving an AI interception by a single aircraft only.

I submit that this information does not constitute strong independent grounds for concluding that the Perkins/BOI-485/Wimbledon scenario is invalidated. The implications are either not clear, or where they are clear they tend to demonstrate an irrelevance to that scenario. It is an incorrect weighting of evidence to subordinate all prior testimony and documentary sources to the recollections of two of these aircrews when there is both documentary and internal evidence of a multiplicity of incidents.

In summary, the reasoning that sees the Perkins/BOI-485 corroboration as mere testimony to confabulation is upside down and relies on a chain of argument that imports assumptions for which there is no evidence. When it was possible to argue that only one pair of jets was sent to intercept one radar-UFO that night it was also possible to argue that inconsistencies between the recollections of Brady/Chambers and Logan/Fraser-Ker and the "classical" Perkins/BOI-485/Wimbledon corpus were damaging to the credibility of the latter. But the evidence now suggests that at least two distinct phases of events, involving a number of attempted interceptions, occurred around Lakenheath that night. The Brady/Chambers and Logan/Fraser-Ker scrambles were associated with one of these phases. In my opinion there are strong arguments for continuing the search for confirmation that the "classical" scenario is an essentially accurate description of an event associated with another phase.

© Martin Shough 2003